International consensus definitions of clinical trial outcomes for kidney failure: 2020

Kidney failure is an important outcome for patients, clinicians, researchers, healthcare systems, payers, and regulators. However, no harmonized international consensus definitions of kidney failure and key surrogates of progression to kidney failure exist specifically for clinical trials. The International Society of Nephrology convened an international multi-stakeholder meeting to develop consensus on this topic. A core group, experienced in design, conduct, and outcome adjudication of clinical trials, developed a database of 64 randomized trials and the 163 included definitions relevant to kidney failure. Using an iterative process, a set of proposed consensus definitions were developed and subsequently vetted by the larger multi-stakeholder group of 83 participants representing 18 different countries. The consensus of the meeting participants was that clinical trial kidney failure outcomes should be comprised of a composite that includes receipt of a kidney transplant, initiation of maintenance dialysis, and death from kidney failure; it may also include outcomes based solely on laboratory measurements of glomerular filtration rate: a sustained low glomerular filtration rate and a sustained percent decline in glomerular filtration rate. Discussion included important considerations, such as (i) recognition of existing nomenclature for kidney failure; (ii) applicability across resource settings; (iii) ease of understanding for all stakeholders; and (iv) avoidance of inappropriate complexity so that the definitions can be used across ranges of populations and trial methodologies. The final definitions reflect the consensus for use in clinical trials. Kidney failure is an important outcome for patients, clinicians, researchers, healthcare systems, payers, and regulators. However, no harmonized international consensus definitions of kidney failure and key surrogates of progression to kidney failure exist specifically for clinical trials. The International Society of Nephrology convened an international multi-stakeholder meeting to develop consensus on this topic. A core group, experienced in design, conduct, and outcome adjudication of clinical trials, developed a database of 64 randomized trials and the 163 included definitions relevant to kidney failure. Using an iterative process, a set of proposed consensus definitions were developed and subsequently vetted by the larger multi-stakeholder group of 83 participants representing 18 different countries. The consensus of the meeting participants was that clinical trial kidney failure outcomes should be comprised of a composite that includes receipt of a kidney transplant, initiation of maintenance dialysis, and death from kidney failure; it may also include outcomes based solely on laboratory measurements of glomerular filtration rate: a sustained low glomerular filtration rate and a sustained percent decline in glomerular filtration rate. Discussion included important considerations, such as (i) recognition of existing nomenclature for kidney failure; (ii) applicability across resource settings; (iii) ease of understanding for all stakeholders; and (iv) avoidance of inappropriate complexity so that the definitions can be used across ranges of populations and trial methodologies. The final definitions reflect the consensus for use in clinical trials. Worldwide estimates suggest that 850 million individuals have kidney disease.1Levin A. Tonelli M. Bonventre J. et al.Global kidney health 2017 and beyond: a roadmap for closing gaps in care, research, and policy.Lancet. 2017; 390: 1888-1917Abstract Full Text Full Text PDF PubMed Scopus (493) Google Scholar,2Jager K.J. Fraser S.D.S. The ascending rank of chronic kidney disease in the global burden of disease study.Nephrol Dial Transplant. 2017; 32: ii121-ii128Crossref PubMed Scopus (99) Google Scholar In 2010, a total of 2.6 million received kidney replacement therapy, and many more died due to lack of access, or other complications such as cardiovascular disease (CVD) and infections.3Liyanage T. Ninomiya T. Jha V. et al.Worldwide access to treatment for end-stage kidney disease: a systematic review.Lancet. 2015; 385: 1975-1982Abstract Full Text Full Text PDF PubMed Scopus (1098) Google Scholar Outcomes for patients with chronic kidney disease (CKD) are comparable to, or worse than, those for patients with other serious diseases, including many forms of cancer.4Heerspink H.J. de Zeeuw D. The kidney in type 2 diabetes therapy.Rev Diabet Stud. 2011; 8: 392-402Crossref PubMed Scopus (62) Google Scholar,5Afkarian M. Sachs M.C. Kestenbaum B. et al.Kidney disease and increased mortality risk in type 2 diabetes.J Am Soc Nephrol. 2013; 24: 302-308Crossref PubMed Scopus (674) Google Scholar Despite the risk, and the wide-ranging impacts of CKD, the numbers of clinical trials in nephrology have lagged behind those in other specialties.6Palmer S.C. Sciancalepore M. Strillpoli G.F. Trial quality in nephrology: how are we measuring up?.Am J Kidney Dis. 2011; 58: 335-337Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar,7Chatzimanouli M.K.T. Wilkens L. Anders H.J. Quantity and reporting quality of kidney research.J Am Soc Nephrol. 2019; 30: 13-22Crossref PubMed Scopus (33) Google Scholar Patients with advanced CKD are often systematically excluded from trials of CVD and diabetes interventions, despite the high CKD prevalence among people with such conditions.8Ishida J.H. Johansen K.L. Exclusion of patients with kidney disease from cardiovascular trials.JAMA Intern Med. 2015; 176: 124-125Crossref Scopus (14) Google Scholar, 9Maini R. Wong D.B. Addison D. et al.Persistent underrepresentation of kidney disease in randomized, controlled trials of cardiovascular disease in the contemporary era.J Am Soc Nephrol. 2018; 29: 2782-2786Crossref PubMed Scopus (34) Google Scholar, 10Zoccali C. Blankestijn P.J. Bruchfeld A. et al.Children of a lesser god: exclusion of chronic kidney disease patients from clinical trials.Nephrol Dial Transplant. 2019; 34: 1112-1114Crossref PubMed Scopus (20) Google Scholar In CKD populations, only a few large trials have demonstrated benefit of interventions.11Baigent C. Herrington W.G. Coresh J. et al.Challenges in conducting clinical trials in nephrology: conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference.Kidney Int. 2017; 92: 297-305Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 12Perkovic V. Jardine M.J. Neal B. et al.Canagliflozin and renal outcomes in type 2 diabetes and nephropathy.N Engl J Med. 2019; 380: 2295-2306Crossref PubMed Scopus (2619) Google Scholar, 13Heerspink H.J.L. Parving H.H. Andress D.L. et al.Atrasentan and renal events in patients with type 2 diabetes and chronic kidney disease (SONAR): a double-blind, randomised, placebo-controlled trial.Lancet. 2019; 393: 1937-1947Abstract Full Text Full Text PDF PubMed Scopus (272) Google Scholar, 14Baigent C. Landray M.J. Reith C. et al.The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection): a randomised placebo-controlled trial.Lancet. 2011; 377: 2181-2192Abstract Full Text Full Text PDF PubMed Scopus (1886) Google Scholar Patients with CKD often have been considered difficult to recruit in numbers required to identify plausible treatment effects, and variability in outcome definitions makes interpretation of and comparison among trials difficult. Reasons for “negative” trials are multifactorial. Recently, the number of nephrology studies in both preclinical and clinical phases has increased, offering the possibility of improving outcomes for patients with CKD.7Chatzimanouli M.K.T. Wilkens L. Anders H.J. Quantity and reporting quality of kidney research.J Am Soc Nephrol. 2019; 30: 13-22Crossref PubMed Scopus (33) Google Scholar With advances in molecular diagnostics and promising new targets for therapies, and efforts to develop new or revisit older therapeutic interventions to tackle kidney diseases and their complications, there is an imperative to enhance the quantity and quality of clinical trials in nephrology worldwide.15Herrington W.G. Staplin N. Haynes R. Kidney disease trials for the 21st century: innovations in design and conduct.Nat Rev Nephrol. 2020; 16: 173-185Crossref PubMed Scopus (12) Google Scholar This need was highlighted in the 2017 “roadmap” for closing the gaps in kidney research, clinical care, and policy.1Levin A. Tonelli M. Bonventre J. et al.Global kidney health 2017 and beyond: a roadmap for closing gaps in care, research, and policy.Lancet. 2017; 390: 1888-1917Abstract Full Text Full Text PDF PubMed Scopus (493) Google Scholar Formalizing the nomenclature of kidney failure for publications has recently been established, after a Kidney Disease: Improving Global Outcomes (KDIGO) Conference.16Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work GroupKDIGO 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease.Kidney Int Suppl. 2013; 3: 1-150Abstract Full Text Full Text PDF Scopus (1578) Google Scholar,17Levey A. Eckardt K. Dorman N. et al.Nomenclature for kidney function and disease: report of a Kidney Disease: Improving Global Outcomes (KDIGO) Consensus Conference.Kidney Int. 2020; 97: 1117-1129Abstract Full Text Full Text PDF PubMed Scopus (258) Google Scholar The Kidney Health Initiative has supported a Clinical Data Interchange Standards Consortium (CDISC), developing data standards for kidney failure outcomes for diabetic kidney disease,18Kidney Health Initiative Clinical Data Interchange Standards ConsortiumDiabetic kidney disease therapeutic area user guide v1.0.www.cdisc.org/standards/therapeutic-areas/diabetic-kidney-diseaseDate accessed: February 16, 2020Google Scholar and extensive analytical work has been conducted with regulators assessing the validity of different surrogates of progression to kidney failure for clinical trials of CKD.19Levey A.S. Inker L.A. Matsushita K. et al.GFR decline as an end point for clinical trials in CKD: a scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration.Am J Kidney Dis. 2014; 64: 821-835Abstract Full Text Full Text PDF PubMed Scopus (352) Google Scholar However, there are no international consensus definitions for use in clinical trials of kidney failure. Over the years, several different definitions for kidney failure have been used, leading to controversy and confusion. The lack of standard definitions has resulted in challenging discussions in adjudication and steering committees, along with regulatory discussions, and resultant varying prevalence of outcomes which impacts power estimations. The absence of universal definitions of kidney failure for trials contrasts with the specialties that have a successful history of conducting large-scale trials and standardized definitions (e.g., myocardial infarction20Thygesen K. Alpert J.S. Jaffe A.S. et al.Fourth universal definition of myocardial infarction (2018).J Am Coll Cardiol. 2018; 72: 2231-2264Crossref PubMed Scopus (1504) Google Scholar and stroke21Sacco R.L. Kasner S.E. Broderick J.P. et al.An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American Heart Association/American Stroke Association.Stroke. 2013; 44: 2064-2089Crossref PubMed Scopus (1820) Google Scholar). Our objective was to develop standardized, internationally accepted, consensus definitions for clinical trial kidney failure outcomes and key surrogates that predict progression to kidney failure. To ensure accountability, transparency, and knowledge translation, the definitions presented were developed with patient partners and a range of stakeholders. This document was reviewed by the collective of meeting participants, and it reflects consensus. The International Society of Nephrology (ISN), through the Research Working Group’s Advancing Clinical Trials (ISN-ACT) Committee, identified the unmet need and facilitated the development of the meeting. A steering committee composed of nephrology and other specialty clinical trialists, ISN leaders, regulatory representatives, and ISN headquarters staff vetted the concept, and made recommendations as to membership of a core group and invitees to the consensus meeting. The stakeholders included patient partners, clinicians and academic clinical trialists, other researchers, regulators, funder representatives, and industry researchers. Invitees were selected based on expertise, and in keeping with the ISN diversity policy (Appendices 1 and 2).22International Society of NephrologyISN position statement on diversity.www.theisn.org/images/Diversity%20Statement/ISN%20Position%20Statement%20on%20Diversity%20%20v3%2022%20Sept%202011.pdfDate accessed: March 9, 2020Google Scholar A total of 105 people from Asia, Africa, Australasia, Europe, Latin and North America, the Middle East, and Russia were invited: 83 attended the consensus meeting. The core group consisted of those with experience steering, designing, and conducting clinical trials, and outcome adjudication, and was chaired by Rajiv Agarwal. Its role was to review the literature of consensus definitions in other domains20Thygesen K. Alpert J.S. Jaffe A.S. et al.Fourth universal definition of myocardial infarction (2018).J Am Coll Cardiol. 2018; 72: 2231-2264Crossref PubMed Scopus (1504) Google Scholar,21Sacco R.L. Kasner S.E. Broderick J.P. et al.An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American Heart Association/American Stroke Association.Stroke. 2013; 44: 2064-2089Crossref PubMed Scopus (1820) Google Scholar,23Hicks K.A. Mahaffey K.W. Mehran R. et al.2017 cardiovascular and stroke endpoint definitions for clinical trials.Circulation. 2018; 137: 961-972Crossref PubMed Scopus (218) Google Scholar,24Leon M.B. Piazza N. Nikolsky E. et al.Standardized endpoint definitions for transcatheter aortic valve implantation clinical trials: a consensus report from the Valve Academic Research Consortium.Eur Heart J. 2011; 32: 205-217Crossref PubMed Scopus (521) Google Scholar and develop a scope-of-work document pertinent to nephrology and kidney failure. Table 119Levey A.S. Inker L.A. Matsushita K. et al.GFR decline as an end point for clinical trials in CKD: a scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration.Am J Kidney Dis. 2014; 64: 821-835Abstract Full Text Full Text PDF PubMed Scopus (352) Google Scholar,25Collins R. Bowman L. Landray M. et al.The magic of randomization versus the myth of real-world evidence.N Engl J Med. 2020; 382: 674-678Crossref PubMed Scopus (202) Google Scholar, 26Pajewski R. Gipson P. Heung M. Predictors of post-hospitalization recovery of renal function among patients with acute kidney injury requiring dialysis.Hemodial Int. 2018; 22: 66-73Crossref PubMed Scopus (29) Google Scholar, 27Siew E.D. Abdel-Kader K. Perkins A.M. et al.Timing of recovery from moderate to severe AKI and the risk for future loss of kidney function.Am J Kidney Dis. 2020; 75: 204-213Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar describes the complete process.Table 1Development steps for the international consensus definitions of clinical trial outcomes for kidney failureStepActivity1Systematically generate a list of clinical trials, including cardiovascular disease, diabetes, kidney disease populations, in which kidney failure was an outcome (see Supplementary Table S1).2Establish a database of definitions of clinical trial endpoints of kidney failure from available trial protocols and/or adjudication manuals (63 randomized controlled trials, 163 definitions; see Supplementary Table S1) and consider publications from the National Kidney Foundation–US Food and Drug Administration–European Medicines Agency (NKF-FDA-EMA) workshops addressing declines in estimated glomerular filtration rate as surrogates for progression to kidney failure for clinical trials of chronic kidney disease.19Levey A.S. Inker L.A. Matsushita K. et al.GFR decline as an end point for clinical trials in CKD: a scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration.Am J Kidney Dis. 2014; 64: 821-835Abstract Full Text Full Text PDF PubMed Scopus (352) Google Scholar,25Collins R. Bowman L. Landray M. et al.The magic of randomization versus the myth of real-world evidence.N Engl J Med. 2020; 382: 674-678Crossref PubMed Scopus (202) Google Scholar, 26Pajewski R. Gipson P. Heung M. Predictors of post-hospitalization recovery of renal function among patients with acute kidney injury requiring dialysis.Hemodial Int. 2018; 22: 66-73Crossref PubMed Scopus (29) Google Scholar, 27Siew E.D. Abdel-Kader K. Perkins A.M. et al.Timing of recovery from moderate to severe AKI and the risk for future loss of kidney function.Am J Kidney Dis. 2020; 75: 204-213Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar3Generate an independent definition of clinical trial outcomes of kidney failure from each core group member, with attention to modifications that may be required in trials conducted in different settings (high- and low-resource settings).4Synthesize a single proposed core group–consensus set of definitions for review at the consensus meeting using an iterative review process.5Develop an accompanying set of questions and controversies for discussion in the consensus meeting breakout groups to ensure scrutiny of the proposed definitions.6Circulate consensus meeting materials to stakeholders 2 weeks before the meeting.7Deliver consensus meeting:(i)Present the need, the remit, and the core group–proposed definitions in plenary sessions, followed by 5 facilitated breakout group discussions.(ii)Present and discuss each breakout group’s key comments in a plenary forum.8Refine definitions based on feedback during the consensus meeting.9Send consensus meeting report and refined definitions for stakeholder comment and public review.10Publish and disseminate. Open table in a new tab Six patient partners participated in the Consensus Meeting, with diverse perspectives, kidney health journeys, and professional interests. They were oriented to the process and the remit during teleconferences, and at the meeting, they voiced their understanding and supported the need of clear definitions for clinical trial kidney failure outcomes. They contributed to all discussions, with their perspectives presented in plenary sessions and incorporated into the definitions. Meeting participants voiced support of concise and clear definitions that could be applied across multiple populations, settings, and a wide range of interventions and trial methodologies. These definitions will support continued efforts to ensure that future clinical trials are comparable, robust, and streamlined.11Baigent C. Herrington W.G. Coresh J. et al.Challenges in conducting clinical trials in nephrology: conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference.Kidney Int. 2017; 92: 297-305Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar,28World Health OrganizationICD-10. International Statistical Classification of Diseases and Related Health Problems. 10th Revision. Vol. 2. 2nd Ed.www.who.int/classifications/icd/ICD-10_2nd_ed_volume2.pdfDate accessed: February 16, 2020Google Scholar Attendees appreciated that irrespective of the final definitions, there will be situations in which modifications are justified or necessary. The consensus was that clinical trial outcomes to represent kidney failure should be comprised of a composite including receipt of a kidney transplant, initiation of maintenance dialysis, and death from kidney failure, and may also include outcomes based solely on laboratory measurements of glomerular filtration rate (GFR)—a sustained low GFR and a sustained percent decline in GFR (Figure 1). Including these separate outcomes should enable kidney failure in various clinical presentations to be well captured in clinical trials. Depending on the trial setting, different components may be included or omitted (e.g., a particularly long and large trial in those with advanced CKD may not need to include the sustained percent decline in GFR, which is a surrogate of progression to kidney failure rather than evidence of kidney failure itself, whereas a trial with less advanced CKD at entry may use that component to ensure a reasonable study duration). Aligned with current clinical classification systems, GFR <15 ml/min per 1.73 m2 is a laboratory-based indicator of kidney failure,16Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work GroupKDIGO 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease.Kidney Int Suppl. 2013; 3: 1-150Abstract Full Text Full Text PDF Scopus (1578) Google Scholar although other thresholds may be used in specific circumstances. We propose the concept of a component called “sustained low GFR,” which permits different thresholds to be used, depending on the population studied and the GFR at study enrollment. The last component is a surrogate, which is sustained percent decline in GFR, highly predictive of progression to kidney failure. Table 2 provides a summary of the definitions for components of the kidney failure composite.Table 2Summary international consensus definitions of clinical trial outcomes for kidney failureComponentsDefinitionKidney transplantationReceipt of a kidney transplantMaintenance dialysisDialysis performed for at least 4 wkDeath from kidney failureThe participant dies, AND kidney replacement therapy was never started (irrespective of reason), AND advanced chronic kidney disease is the underlying cause of deathSustained low GFRGFR <15aA GFR of <10 ml/min per 1.73 m2 may be used in certain situations. ml/min per 1.73 m2 sustained over at least 4 wkSustained percent decline in GFRPercent decline in GFR of ≥40%bA ≥30%, ≥40%, ≥50%, or ≥57% decline in GFR (a 57% decline in estimated GFR approximately corresponds to a doubling of serum creatinine) may be considered surrogates for progression to kidney failure depending on the trial population and acute effects of study intervention on GFR. A doubling of creatinine is the most well established, whereas a ≥30% decline, in comparison, is the least reliable of these (putative) surrogates. from a baseline start point sustained over at least 4 wk19Levey A.S. Inker L.A. Matsushita K. et al.GFR decline as an end point for clinical trials in CKD: a scientific workshop sponsored by the National Kidney Foundation and the US Food and Drug Administration.Am J Kidney Dis. 2014; 64: 821-835Abstract Full Text Full Text PDF PubMed Scopus (352) Google ScholarGFR, glomerular filtration rate.a A GFR of <10 ml/min per 1.73 m2 may be used in certain situations.b A ≥30%, ≥40%, ≥50%, or ≥57% decline in GFR (a 57% decline in estimated GFR approximately corresponds to a doubling of serum creatinine) may be considered surrogates for progression to kidney failure depending on the trial population and acute effects of study intervention on GFR. A doubling of creatinine is the most well established, whereas a ≥30% decline, in comparison, is the least reliable of these (putative) surrogates. Open table in a new tab GFR, glomerular filtration rate. Kidney transplantation is defined as receiving a kidney transplant irrespective of source (cadaveric vs. living donor), or successful implantation or graft function. The date of the procedure constitutes the outcome’s date. No modifications are required for low- or high-resource settings. Maintenance dialysis is defined as dialysis (peritoneal or hemodialysis) performed for at least 4 weeks. A number of different confirmatory periods have been used (e.g., 4 weeks; 30, 60, or 90 days) in different trials; we agreed on 4 weeks, for the sake of harmonization, after careful discussion. There are situations in which the duration of at least 4 weeks may not be met, for example, due to death, or further data are unavailable, but an outcome of kidney failure requiring dialysis can be inferred. These situations include start of dialysis after a trajectory of progressive CKD, discontinuation of dialysis treatment within a few days due to either futility of therapy (dialysis withdrawal) or resource constraints, or kidney transplantation shortly after dialysis initiation. All of these situations constitute reaching a trial outcome of maintenance dialysis (see section on adjudication below). One point that was highlighted is that although at least 4 weeks of dialysis is recommended for clinical trials, a longer duration (e.g., 90 days) is in keeping with most “registry” definitions of maintenance dialysis. We also acknowledged that the duration of dialysis that confirms it to be maintenance dialysis may differ by the specific patient population under study (e.g., patients with vasculitis or glomerulonephritis may justify a longer duration). Although some patients may recover from dialysis after being on it for 4 weeks,29Levey A.S. Stevens L.A. Schmid C.H. et al.A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150: 604-612Crossref PubMed Scopus (15959) Google Scholar practical considerations of timely trial completion, and most importantly, the patient perspective that “any dialysis duration constitutes kidney failure” was appreciated. Thus, we agreed that 4 weeks was a suitable definition in most circumstances and would appropriately exclude most recoverable dialysis-requiring acute kidney injury (AKI) from a maintenance dialysis outcome. Those who recover dialysis-independent kidney function could have this outcome revised and be included in a sensitivity analysis to determine the impact of that change.30Greene T. Teng C.C. Inker L.A. et al.Utility and validity of estimated GFR-based surrogate time-to-event end points in CKD: a simulation study.Am J Kidney Dis. 2014; 64: 867-879Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar From the patient perspective, any duration of dialysis is a serious failure of their kidneys. Therefore, trials that report kidney failure as an outcome should collect and report information on AKI requiring dialysis or consider reporting a patient-centered outcome of “any dialysis.” In low-resource settings, no modifications to the definition are required; however, it was recognized that participants on maintenance dialysis may be dialyzed less frequently than those in high-income countries and may discontinue therapy because of non-availability due to financial or other reasons. Dialysis frequency is thus not included in the outcome’s definition. This component usually represents a small proportion of kidney failure outcomes. Its inclusion allows recognition that those participants who progressed to a GFR <15 ml/min per 1.73 m2, and often one much lower, and died as a result of non-availability of kidney replacement therapy or a personal decision to not pursue dialysis, have reached the outcome. This is analogous to the KDIGO nomenclature of death from “kidney failure without replacement therapy.” In these situations, the underlying cause of death (i.e., the disease that initiated the train of morbid events leading directly to death31Coresh J. Turin T.C. Matsushita K. et al.Decline in estimated glomerular filtration rate and subsequent risk of end-stage renal disease and mortality.JAMA. 2014; 311: 2518-2531Crossref PubMed Scopus (639) Google Scholar) is advanced CKD. In the context of trials, distinguishing between death with kidney failure and death from kidney failure is important. The death from kidney failure concept presumes that the driver of death is the low kidney function (irrespective of specific mechanisms, such as hyperkalemia, fluid overload, etc.). There is no need for modification of this definition for different resource settings, but we recognized that death from kidney failure may be more common in lower-resource settings. It is noted that those who develop AKI with no pre-existing CKD, and who either are not offered dialysis or refuse it, or for whom it is unavailable, and subsequently die, may or may not be considered an instance of death from kidney failure, depending on circumstances, and study design. In the conduct of the trial, this should be prespecified. Meeting attendees deliberated on the concepts of death from kidney failure, and death with kidney failure. There was extensive debate, but there was some agreement that death with kidney failure should not be recommended to be a standard component of the kidney failure outcome definition (Figure 1). There is a need to differentiate deaths from other causes (e.g., overwhelming sepsis or fatal CVD) in people with kidney failure, from the deaths of those who die from kidney failure (due to non-provision of replacement therapy for any reason AND advanced CKD). We appreciate the sometimes subtle differences and recommend that death with kidney failure be considered as a subsidiary outcome in clinical trials. Death with kidney failure includes death, irrespective of cause, among those who have a clinical outcome of maintenance dialysis, kidney transplantation, or a sustained low GFR (but, importantly, not a sustained percent decline in GFR alone). Because an increasingly high proportion of patients with CKD in developed countries are elderly and multimorbid, death with kidney failure may become more common. GFR-based outcomes represent nonclinical laboratory outcomes that are acceptable in clinical practice and trials. Some laboratory-based outcomes have been shown to strongly predict progression to clinical outcomes